DMM probes plugged at mains while NOT in ACV measuring mode

[BravoV]

We're not talking cheap or dodgy DMM here, does good 1st tier DMM "by design" will survive when it's probes on mains while its NOT in AC Volt measurement mode, like in DC (V or miliVolt) mode, Ohm (includes continuity mode), Diode test, Capacitance test, Temperature test ?

Obviously this question excludes Current measurement mode.   

[Wytnucls]

Most do survive, even some of the cheap ones.
This one didn't like mains on the mA/uA polyswitch though, giving up the ghost on that setting. 

[sacherjj]

A properly designed DMM can survive main connection in any mode with possible exception of current, as you mention.  It is likely that current modes may fuse the proper fuse, without hurting the rest of the meter.  With slow blow 10A style, that might not work over 120/220V.

[electronics man]

[BravoV]

Yes, I've seen that particular Dave's video, from that example its obvious the Agilent meter design is not the best one, isn't it ? cmiiw

Its just I'm expecting maybe some of you have studied or better experienced working on the input circuit that aware "by design" of a good DMM, please share your insight, especially on each test modes, not only just at Ohm mode as Dave did.

[Wytnucls]

On some meters, I have seen what I think is a crowbar device to protect some ranges from overvoltage:

[Monkeh]

On some meters, I have seen what I think is a crowbar device to protect some ranges from overvoltage:

Diode clamping, and the current is interrupted by a fuse or limited by the PTC. You'll find bridge rectifiers and/or discrete diodes on the mA range in Fluke meters to do this:

[Wytnucls]

I'm familiar with diode clamping on the mA range, but don't really understand how the clamping is done on the Ohms range with the 2 NPN transistors. Someone would care to explain?

http://www.fairchildsemi.com/ds/SS/SS8050.pdf

[Wytnucls]

It looks like the transistors are reversed biased during an overvoltage, resulting in a breakdown voltage emitter-base of 6V.

[drtaylor]

DMM protection circuits are built around current limiting and voltage clamping. A well designed DMM will easily shrug off line voltage in the Volt and Ohms input. In the voltage input a current limiting resistor is followed by a voltage clamp. The voltage clamp for high voltage has been traditionally series connected varistors after a power resistor. The power resistor of choice was usually a carbon composition. Carbon comps are fairly crappy resistors except when it comes to pulse and instantaneous power surges. In many DMMs the protection for direct inputs tend to be multiple resistors as high as 220k and several watt ratings. In the early Fluke DMMs usually a 220k 1W Carbon Comp resistor was the main overvoltage protection. The input clamps of the ICs were all that was necessary as a 250VAC line input would clamp at 5V. Thus the current was limited to 240/220k = ~1.1mA. The power dissipated by the resistor would be far less than 1 W. Protecting the ohms input takes a little more finesse. The back to back diode connected transistors provides a low leakage clamp at about ±6.7 volts. If a main supply appears the clamp conducts. It has to sustain the energy long enough for a PTC thermistor to react. Once the PTC has heated up, the current into the current source is reduced to a very low level. The PTC thermistor are only used in the current source leads, as they have too much noise on their own. In later DMMs the clamp got more sophisticated using voltage distribution to ensure that the ohms circuit would survive even 1000V overloads. The PTC thermistor recovery time is why some DMMs require a while to work correctly in the ohms settings after being hit with a large overvoltage.

Once you are past the lowest volts settings, the inherent current limiting of the input divider kicks in. Since the first resistor in the divider is usually 9Meg ohms not much current is going to flow.

The real challenge in protecting DMM inputs is to have effective clamping that does not leak current, or add excess noise to the desired measurement. I could go on and on about various methods of low leakage clamping and other circuit protection, but it will take more time than I have right now.

[Napalm2002]

Thanks for the info Mr Taylor. Very cool info!

[AG6QR]

A well designed meter will survive mains voltage on any setting, perhaps with a blown fuse, but no other harm done.

My first Fluke, a model 16 (a very basic and now obsolete meter intended for HVAC technicians) will not only survive, but if it's in ohms, continuity or diode check mode, it will automatically switch to lo-Z voltage mode and display the measured voltage.  It does this any time it senses a voltage greater than about 4.5V.  It will also automatically figure out whether the voltage is AC or DC.

In fact, that's the only way to access the Lo-Z voltage mode on that meter.  Put it in ohms (or continuity or diode check) mode and connect the probes to a voltage.

There are some other more modern Flukes that have a similar feature, but it's found mostly on the "service tech" line of meters, and not the high precision meters for electronics work.  The Fluke 113 takes it to an extreme, with no dedicated ACV or DCV setting, but a single setting that does ACV, DCV, continuity, and diode check, depending on what it senses at its terminals.  That meter is designed for utility workers, and it looks to me like the chief design goals were ruggedness and idiot-proofing for the kinds of measurements a utility worker needs to do.

[Wytnucls]

The Gossen MetraHit 28S/29S both have an auto detection selection on the range switch.
That range selection will auto detect, after 1 second, AC and DC voltages up to 500V, resistance up to 15MOhm, capacitance between 1.5nF and 300uF and can also measure diodes with a current of 1V.
They are both CAT III 1000V/CAT IV 600V, reduced later to CAT III 600V/ CAT IV 300V.

[retiredcaps]

I could go on and on about various methods of low leakage clamping and other circuit protection, but it will take more time than I have right now.

Thanks for adding your unique insight into this thread.  I look forward to hearing more when you have time. 

I'm trying to learn about all aspects of the multimeter and find this all very instructive.

[GiskardReventlov]

I wonder why probes aren't designed to fail first before allowing the meter to be fried. Is it a safety issue? Melting cable slag could be painful. Or would cables with fuses alter the measurements?

[HKJ]

I wonder why probes aren't designed to fail first before allowing the meter to be fried. Is it a safety issue? Melting cable slag could be painful. Or would cables with fuses alter the measurements?

When the DMM reads 0 how would you know if you have a de-energized circuit or a probe with blown fuse?

[BravoV]

I could go on and on about various methods of low leakage clamping and other circuit protection, but it will take more time than I have right now.

Thanks for adding your unique insight into this thread.  I look forward to hearing more when you have time. 

I'm trying to learn about all aspects of the multimeter and find this all very instructive.

+1 fully agree. 

I guess I just opened a can of worm on this specific topic as Dave loves to say. 

Thanks a lot drtaylor ! 

[AG6QR]

I wonder why probes aren't designed to fail first before allowing the meter to be fried. Is it a safety issue? Melting cable slag could be painful. Or would cables with fuses alter the measurements?

Poorly designed meters might fail merely due to high voltage with very little current.  Probes can't stop high voltage while passing low voltage, because the high voltage is present between one probe and the other, while there's no significant voltage or current between a probe's tip and the banana plug.  So if voltage protection is needed, it's got to go in a place where the protection can see both probes, i.e., it's got to go in the meter.

Current protection is another story.  Fused probes are certainly possible.  But there's a human factors issue with putting fuses in the probes:  if a fuse is blown, the probe will probably lose continuity and show 0 volts.  This might lead a user to believe a circuit was off when in fact it was energized with dangerous voltage.  There's also a couple of technical issues: if you're going to have a fuse blow and dissipate high energy, it's easier to contain the energy in a bulky meter than in a slim probe.   Finally, it's hard to make different levels of fuses for different current ranges in probes, while insuring that the user is using the correct fuse for the selected range.

There are some ESA guidelines that call for probes to be protected by fuses.  But they're not without controversy, and some people have suggested that unfused, with a well-protected meter, is at least as safe, and maybe safer. 

http://www.fluke.com/fluke/caen/accessories/Test-Leads-Probes-and-Clips/FTP-1.htm?PID=73763

[GiskardReventlov]

Current protection is another story.  Fused probes are certainly possible.  But there's a human factors issue with putting fuses in the probes:  if a fuse is blown, the probe will probably lose continuity and show 0 volts.  This might lead a user to believe a circuit was off when in fact it was energized with dangerous voltage.

This was mentioned by HKJ too.  Could a sense circuit on the meter detect the blown probe fuse? Then the meter just beeps and does nothing else.

I think we've found a use for that robotic arm that Dave got.  High current probe and meter testing time!

Fluke solved it like so:
"Voltage readings with blown fuse are approximate and vary with the meter impedance. Before and after each measurement, verify presence of blown fuse by checking continuity of the fused test probe. Current mode is non-functional with blown fuse."


Yes I would rather fry the meter than my hand but if the meter fries then the probe tips are probably molten slag also.
It's an interesting problem. Maybe the best weapon is to remember: don't stick your probe into dangerous places.
Better to play it safe and live to probe another day.

[electronics man]

mabe if you dont plan on using very high voltages that what about a polyswitch in the probe

[drtaylor]

Putting a Polyswitch in a probe is a very bad idea, even if you limit the voltage to what the polyswitch is rated. The nature of polyswitches guarantee that they are a large noise source. It would compromise your performance at low voltages. Raychem does have one rated to 600VAC. But you put 1200V across that and blooieeee! You cannot get more voltage by putting them in series because they'd never balance properly. Anyway bottom line is it is not a safe or good way to protect a DMM. Polyswitches are good for protecting current source outputs used for ohms as the voltage that is measured never includes the Polyswitch. Plus you can clamp the voltage well below the polyswitch rating.  The same goes for Ratiometric ohmmeters found in most dual slope DMMs.